llama-exp

Experiments with AI using llama index library

This is a very simple setup of using llama-index to ingest data as index and to query your data using OpenAI as LLM.

Notes (on data):

• You can change you input data which can be indexed and queried upon
• The program is using UnstructuredReader which reads unstructured data and converts it into text document
• You can change the data files in backend/data/uploaded folder and update SOURCE_FILE_PATH in the ingest_data.py to your file path

Pre-requisite

1. You should have OPENAI_API_KEY in your environment, if you already do not have the key you can setup an account OpenAI
2. Create a .env file in the backend folder and add the OPENAI_API_KEY to the file

OPENAI_API_KEY=<your_key>

Setup :

1. Create a python virtualenv on your system

virtualenv -p python3 llama-env

# Activate the environment
source ./llama-env/bin/activate

2. Install depdencies

pip install -r requirements.txt

3. Ingest data

cd backend
python src/ingest_data.py

4. Query data

cd backend
python src/query.py What is this document about?